HOSTED BY Available online at www.sciencedirect.

com

ScienceDirect
Soils and Foundations 59 (2019) 942–954
www.elsevier.com/locate/sandf

Technical Paper

Strength of sandy and clayey soils cemented with single and double
fluid jet grouting
Modoni Giuseppe a,⇑, Wanik Lidia b, Mascolo Maria Cristina a, Salvatore Erminio a
Bzówka Joanna b, Shen Shui-Long c, Daniele Valeria d, Pingue Luca e
a
University of Cassino and Southern Lazio, via di Biasio 43, 03043 Cassino, FR, Italy
b
Silesian University of Technology, Akademicka 5, 44-100 Gliwice, Poland
c
Jiao Tong University, 800 Dong Chuan Rd., Shanghai 200240, China
d
University of L’Aquila, Piazzale E. Pontieri 1, 67100 Monteluco di Roio, L’Aquila, AQ, Italy
e
TREVI s.p.a., Via Dismano, 5819, 47522 Cesena, Italy

Received 3 October 2018; received in revised form 15 March 2019; accepted 28 March 2019
Available online 30 May 2019

Abstract

Innovations in jet grouting technology have primarily focused on the cutting efficiency of the jets, with the aim of creating larger columns
and increasing the productivity of construction sites. Relatively little attention has been paid to the consequences of the grouting system on
the mechanical properties of the formed material. This paper investigates this aspect by analysing the results of two field trials carried out in
both sandy and clayey soils, where single and double fluid jet grouting were simultaneously performed, with varied grout composition and
injection parameters. Parallel uniaxial compressive tests on samples cored from the columns show that the material formed with the double
system is systematically lower in strength than the material formed using the single fluid system. The mineralogical composition of samples
cored from the columns was analysed by performing parallel Scanning Electron Microscopy (SEM), X-ray diffraction analysis (XRD), Dif-
ferential Thermal Analysis (DTA) and Thermo-Gravimetric Analyses (TGA) to determine the reasons for this difference. A lower proportion
of cementitious products, an accelerated carbonation of portlandite and a less homogeneous distribution of cement hydration products was
found on the surface of the soil particles of the double samples than for the single fluid columns.
Ó 2019 Production and hosting by Elsevier B.V. on behalf of The Japanese Geotechnical Society.

Keywords: Jet grouting; Uniaxial compressive strength; Sand; Clay; Mineralogy

1. Introduction geotechnical applications ranging from foundation rein-

forcement (Modoni and Bzόwka, 2012), impervious walls
From its earliest application (e.g. Miki, 1973), jet grout- (Croce and Modoni, 2007), bottom plugs (Modoni et al.,
ing has greatly progressed (e.g. Croce et al., 2014) and now 2016a), and tunnels (Ochmański et al., 2015a; Qiu et al.,
competes with other engineering solutions in many 2018; Atangana Njock et al., 2018a). The basic principle
of the technology is to form columns of cemented material
in situ, mixing the soil with high speed jets of water-cement
Peer review under responsibility of The Japanese Geotechnical Society. grout. With time, the challenge became to increase the
⇑ Corresponding author.
pumping power and the hydrodynamic efficiency of the
E-mail addresses: [email protected] (G. Modoni), [email protected]
(L. Wanik), [email protected] (C. Mascolo Maria), e.salvatore@ injection systems in order to make the technology more
unicas.it (E. Salvatore), [email protected] (J. Bzówka), slshen@sjtu. productive, increasing the size of columns and conse-
edu.cn (S. Shui-Long), [email protected] (V. Daniele), lpingue@ quently reduce the time required for their installation.
trevispa.com (L. Pingue).

https://doi.org/10.1016/j.sandf.2019.03.007
0038-0806/Ó 2019 Production and hosting by Elsevier B.V. on behalf of The Japanese Geotechnical Society.
 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954 943

Nowadays, columns with diameters as large as three meters control of the mechanical properties of the material based
and more are possible (Eramo et al., 2012). In addition, the on a deeper knowledge of their relationship with the exec-
most recent equipment makes it possible to selectively con- utive factors may increase confidence in this technology,
trol the rotating speed of the monitor and create columns optimize its use and, perhaps, pave the way to assigning
of non-circular sections (e.g. elliptical, rectangular, square), a permanent function to jet grouted structures.
which is particularly convenient when irregular volumes Studies performed on samples of cement stabilised
must be covered (e.g. Flora et al., 2017). Notably, the inno- clayey soil carefully reconstituted in the laboratory with
vations have resulted in a marked reduction in the number variable composition (e.g. Lee et al., 2005; Kitazume and
of boreholes and in the time needed to create massive struc- Terashi, 2013; Atangana Njock et al., 2018b) contribute
tures like bottom plugs or cut-off walls. The larger effi- to quantify the influence of constituent materials (water,
ciency of the cutting systems has considerable cement and soil) and their relative proportions on the
environmental benefits as well, since a lower amount of strength of cemented soil. Atangana Njock et al. (2018b)
spoil is produced. In this race, one of the most effective noted a noticeable reduction of the uniaxial compressive
technological innovations is represented by the so-called strength with the water-cement ratio of the mix, confirming
double-fluid jet grouting. This injection system consists of previous results from Kaushinger et al. (1992). Lee et al.
two coaxial jets, an internal jet of massive fluid, which aims (2005) also found a slight decrease in the cement-soil ratio.
at eroding the soil, an outer annular jet of air which func- This apparently surprising result is motivated by the
tions to coat the inner one and protect it against the turbu- authors with a more flocculated and porous structure of
lent exchange of forces and mass with the material filling the clayey material containing a higher fraction of cement.
the space outside the nozzle (Modoni et al., 2006; 2016b; Kitazume and Terashi (2013) have provided a compre-
Flora et al., 2013, Ochmański et al., 2015b, Shen et al., hensive classification of the factors influencing the chemical
2013). This principle is also at the base of the triple fluid reaction between soil and binder (here reported in Table 1).
system, where an air-assisted water jet is firstly injected to For a complete overview of the dependency of the uniaxial
cut the soil and a second jet of cement grout follows to pro- compressive strength on each factor, readers may refer to
vide cementation. Nowadays, the possibility of getting their original publication. It must, however, be pointed
higher pumping pressures and flow rates has made the dou- out that the above studies consider clayey soils, and the
ble fluid system the most popular, with applied energies as results therefore cannot be extended to non-plastic materi-
high as 1400% above the first application (see Fig. 1). This als (e.g. sand) treated with jet grouting.
system is considered effective in almost all soil types, even In considering the objective of the present study, i.e. the
those particularly resistant to cutting. role of jet grouting injection system, many of the factors
While there is much concern about the dimensions of listed in Table 1 are independent on the technology. Those
columns, little attention is paid to the consequence of the which might vary between single and double fluid injection
injection system on the mechanical properties of the jet- systems, and thus deserve to be carefully analysed, have
grouted material. This situation basically is basically due been marked with an asterisk. Even when the injection
to the overabundanc of jet grouted structures and their parameters are assigned and measured, the composition
provisional functions. That is, high strength is generally of jet grouting material cannot be known with a good accu-
not a paramount requirement. Nonetheless, more accurate racy, and some uncertainty will always exist. This is

Fig. 1. Time evolution in the use of jet grouting (from Flora et al., 2017).
944 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954

Table 1
Factors affecting strength increase of cement stabilised soil (derived from Kitazume and Terashi, 2013).
I. Characteristics of binder 1. Type of binder
2. Quality
3. 3. Mixing water and additives (*)
II. Characteristics and conditions of soil 1. Physical, chemical and mineralogical properties of soil (especially for clays)
2. Organic content
3. Potential Hydrogen (pH) of pore water
4. 4. Water content
III. Mixing conditions 1. Degree of mixing (*)
2. Timing of mixing/re-mixing
3. 3. Quantity of binder (*)
IV. Curing conditions 1. Temperature
2. Curing period
3. Humidity
4. Wetting and drying/freezing and thawing, etc.
5. Overburden pressure
(*) Parameters likely to be influenced by the jet grouting injection system

because the formation of the column is ruled by factors systematic collection and interpretation with advanced sta-
controllable up to a limited extent, primarily the variability tistical methods (Support Vector Machines) of field data
in the soil composition and the imperfect mixing with from different jet grouting projects. Considering all the
grout. One of the most evident effects is the considerable possible factors, these authors found the injection system
heterogeneity of the material that results in variation coef- to have the second highest relative importance on the
ficients of the uniaxial compressive strength as high as 70% unconfined compressive strength of the produced material,
(Toraldo et al., 2018). However, despite this variability, the the first being the ratio between porosity and cement con-
few studies attempting to quantify the role of the injection tent of the material. Considering also that porosity and
system (e.g. Van der Stoel, 2001) have reported a systemat- cement content of the material are both influenced by the
ically lower strength of material created with the double injection system, it is evident that the technology deserves
fluid system than that created with the single fluid to be investigated with more attention.
(Fig. 2). A strong dependency on the injection system is Starting from these premises, the present study inter-
also envisaged by Tinoco et al. (2014) who carried out a prets the influence of the injection system on the mechani-
cal properties of jet grouted material from a
micromechanical perspective. To this aim, classical
mechanical tests have been coupled with microscopic
observation and thermo-physical tests to investigate the
mineralogical composition of samples formed in two differ-
ent soil types (cohesive and cohesionless) with single and
double fluid jet grouting.

2. Experimental study

To investigate the dependence of the injection system on

the strength of the jet grouted material, two field trials have
been performed. In both cases, columns were simultane-
ously created with single and double fluid jet grouting sys-
tems, then their top portions were excavated, and samples
were core, taken to the laboratory and investigated with a
set of mechanical and chemical tests. Considering that
the properties of the jet material depend significantly on
the constituent soil (e.g. Toraldo et al., 2018), two test
sites were selected for the evaluation of the performance
of the injection systems in both sandy and clayey
deposits.

2.1. Case study #1: Sandy soils

Fig. 2. Strength of jet grouted columns created in the same soil with single This first field trial was performed in Silesia (Poland)
and double fluid jet grouting (adapted from Van der Stoel, 2001). with the general purpose of investigating the effects of
 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954 945

Fig. 3. Field trial in sandy soils: stratigraphy of the subsoil (a) and layout of columns (b) (Wanik, 2017).

injection systems and parameters on the diameter and tests performed on specimens with a height to diameter
mechanical properties of columns (Wanik, 2017; Wanik ratio (h/d) equal to one or two give comparable results.
et al., 2017). The subsoil stratigraphy was investigated with On the other hand, there is a very strong influence on the
two boreholes and four Piezocone Tests (CPTU) to deter- injection systems as the measured Uniaxial Compressive
mine the alluvial sequence shown in Fig. 3a, with a top Strength (qu) range between 11.7 and 11.8 MPa for the sin-
layer of medium sand superposed to organic clays and gle fluid material, between 1.1 and 3.2 MPa for the double
coarse sand. Sixteen columns of about 4 m in length (from fluid system, i.e. the strength is from 365 to 1000% higher
about 0.5 m to 4.5 m below ground level) were created, for the single than for the double fluid. Such a difference
eight with single fluid and eight with double fluid system cannot be explained by the higher presence of voids in
(indicated respectively as 1S–8S and 1D–8D in Fig. 3b). the double fluid material (induced in principle by the injec-
The same lifting speed was assigned to all columns, while tion of air), as the dry unit weight of the two materials (sin-
the flow rate was modified by parametrically modifying gle and double fluid) is similar (see Table 3).
the diameter of the nozzles and injection pressure (Table 2). Therefore, an attempt has been made to correlate the
The same binder type, Portland composite cement mixed variation of strength seen in Table 2 with the composition
with fly ash (code CEM II/B-V 32.5 R - EN 197-1, 2011), of the materials with particular attention to the amount of
and cement-water (equal to 1) were y used for all columns. cement in relation to water and soil. In fact, as the adopted
Columns were then excavated up to 1.5 m depth below parameters (namely grout composition and flow rates, lift-
the ground surface (Fig. 4a) to measure the diameters ing speed of the monitor) are similar for the two injection
and to take core samples representative of the material systems and the diameter larger for the double fluid col-
(see Fig. 4b). umns, a lower presence of cement must be assumed for
The mean diameters of the single and double fluid jet the second case. To quantify this effect, the cement content
grouting columns, reported in Table 2 together with the in each column has been calculated starting from the com-
injection parameters, show a clear dependency on the grout position of grout (cement–water ratio was equal to 1 in all
flow rate. Although important, this issue is out of the scope cases), the injected flow rate, the lifting speed of the mon-
of the present work. Here, it is just important to note that itor and the diameter of the columns (Table 2). Water-
the columns created with double fluid had larger dimen- cement and cement-soil ratios have been estimated assum-
sions than those formed with the single fluid system. ing an initial porosity of the soil equal to 0.3 and equal
composition for the material present in the column and
2.1.1. Uniaxial compressive tests in the spoil. It should be noted that because it is not possi-
Several samples were extracted from the top portion of ble to retrieve and quantify the amount of cement directly
the columns, i.e. the one formed in the medium sand layer from samples, indirect estimates were made. These esti-
(see Fig. 4b), and were taken to the laboratory and sub- mates affect the absolute values of the quantities computed
jected to uniaxial compression tests. The results, also in Table 2, but certainly do not alter the relative conditions
reported in Table 2, show the differences between the mate- between single and double fluid material.
rials created with single and double fluid. The statistical The computed values are reported in Table 2 and plot-
analysis of these data, summarized in Table 3, shows some ted againt uniaxial compressive strength in Fig. 5. Despite
limited relevance to the slenderness of the samples, as the some unavoidable scattering, the position of the two
 946 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954

families of dots reveals a clear correlation between the

adopted variables, and their relative position justifies the
compressive strength

different uniaxial compression strength. The dependency

on the water-cement ratio is in complete accordance with
Mean uniaxial

the results provided by Atangana Njock et al. (2018a,

2018b) and Lee et al. (2005), while the dependency on the
cement-soil ratio follows an opposite trend. However, it
[Mpa]

20.80

11.60
11.08
11.31
11.07
7.80
4.55

1.60
3.40
3.10
3.30
1.60
5.30
1.30
must not be forgotten that the present tests have been con-
ducted in sandy soils (Iolli et al., 2015; Salvatore et al.,
2017), whereas Lee et al. (2005) refer to clayey soils which
cement ratio

are characterized by different chemical reactions with

Water-

cement. Fig. 5b shows that the uniaxial compressive

1.63
1.89
1.57
1.74
1.63
1.97
1.57
1.90
2.52
2.82
2.48
2.75
3.39

4.57
4.22
strength increases significantly with the cement/soil ratio
and this result suggests it is possible to accurately control
Cement-soil

the amount of injected cement considering the size of col-

umns produced with each technology.
ratio

0.61
0.43
0.68
0.52
0.61
0.40
0.68
0.43
0.25
0.21
0.26
0.22
0.16

0.11
0.12

2.1.2. Mineralogical composition

The role of the injection system was investigated by a
series of laboratory tests, using small pieces of material
Injection parameters and fundamental properties of the different columns of the field trial performed in sandy soils (Wanik, 2017).

from the tested samples to get a deeper insight in the min-

volume of column

eralogical composition of the two materials. An immediate

Cement per unit

confirmation of the different presence of cement in the

material for single and double fluid system is seen from
[kg/m3]

the Scanning Electron Microscope (SEM, Reimer, 1998)

291
232
310
264
291
219
310
231
157
135
159
139
107

analysis, whose typical results are reported in Fig. 6. In

76
83

fact, the sample treated with single fluid (Fig. 6a) shows
diameter [m]

a diffused presence of plate-like portlandite crystals aggre-

gates (Mascolo et al., 2010) and calcium silicate hydrate
Mean

particles (CSH) surrounding the bigger quartz grains.

1.13
0.94

0.78
1.13
0.98

0.86
1.75
1.61
1.73
1.58
1.54

1.88
1.63
0.9

0.9

These minerals, scarcely present in the double fluid mate-

–

rial (Fig. 6b), are produced from the hydration of cement

Lifting speed

and are the main reason for the binding action, and there-
fore, the biggest contributors to the material strength
0.0083
0.0083
0.0083
0.0083
0.0083
0.0083
0.0083
0.0083
0.0083
0.0083
0.0083
0.0083
0.0083
0.0083
0.0083
0.0083
[m/s]

(Taylor, 1997).
Along with natural soil directly taken from the test site,
the two samples of material created with single and double
fluid were subjected to X-ray diffraction (XRD, Klug and
Grout flow
rate [m3/s]

Alexander, 1974), differential thermal (DTA, MacKenzie,

0.0053
0.0026
0.0037
0.0022
0.0053
0.0026
0.0037
0.0022
0.0053
0.0037
0.0053
0.0037
0.0026
0.0022
0.0026
0.0022

1970) and thermogravimetric analyses (TGA, Coats and

Redfern, 1963). The XRD technique allowed the crystalline
composition of the materials to be characterized. The
typical pattern (Fig. 7) shows several peaks, each produced
by the constructive interference of a monochromatic beam
pressure

of X-rays diffracted at specific angles from a set of lattice

Grout

[MPa]

planes in the crystalline structure of the material. The

36
36
18
25
36
36
18
25
36
18
36
18
36
26
36
25

intensities of each peak are determined by the distribution

of atoms within the crystalline lattice. Consequently, the
X-ray diffraction pattern is the fingerprint of the periodic
Nr and diameter
[mm] of nozzles

atomic arrangements in a given material structure. To iden-

tify the different components, the observed diffraction
peaks are compared with an official database of the Inter-
2x4.0
2x2.8
2x4.0
2x2.8
2x4.0
2x2.8
2x4.0
2x2.8
2x4.0
2x4.0
2x4.0
2x4.0
2x2.8
2x2.8
2x2.8
2x2.8

national Centre for Diffraction Data (ICDD, 2017) associ-

ating each angle to a specific crystalline phase.
The XRD pattern of the natural soil (Fig. 7a) reveals
Column
Table 2

that quartz, SiO2, (JPCDS card no. 46-1045) is the preva-

1D
2D
3D
4D
5D
6D
7D
8D
1S
2S
3S
4S
5S
6S
7S
8S

lent crystalline phase of the sand and that dolomite

#
 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954 947

Fig. 4. Top of the excavated columns (a) and cored samples (b) (Wanik, 2017).

Table 3
Statistics of the different laboratory tests performed on the material coming from the field trial in sandy soils.
Shape ratio h/d h/d = 1.0 h/d = 2.0
Jet grouting system Single Double Single Double
Number of samples 22.0 15.0 19.0 8.0
Mean dry unit weight [kN/m3] 16.5 16.9 16.9 17.4
Standard deviation unit weight [kN/m3] 1.2 1.0 1.9 0.5
Mean qu [MPa] 11.7 3.2 11.8 1.1
Standard deviation qu [MPa] 1.6 0.5 1.1 0.1

Fig. 5. Dependency of the uniaxial compressive strength on the estimated water-cement (a) and cement-soil (b) ratios of the jet grouted material.

Fig. 6. SEM micrographs of samples treated with single (a) or double (b) fluid system.
948 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954

Fig. 7. XRD patterns of the natural soil (a), of the soil treated with single (b) and double (c) fluid jet grouting.

ferroan, Ca(Mg, Fe)(CO3)2, (JPCDS card no. 34-517), change of weight while samples are exposed to progres-
muscovite, KAl3Si3O10(OH)2, (JPCDS card no. 1-1098) sively increasing temperatures. The latter test, run simulta-
and microcline, KAlSi3O8, (JPCDS card no. 19-932), are neously with the previous one, measures the difference in
present with progressively lower amounts. The same test temperature between the sample and an inert reference
performed on soil treated with single fluid jet grouting material while the temperature of the room is increased.
(Fig. 7b) reveals the presence of the above components The peaks of this variable indicate that a chemical-
together with portlandite Ca(OH)2, (JPCDS card no. physical transformation, whether endothermic or exother-
44-1481), calcite, CaCO3, (JPCDS card no. 13-192), tetra- mic, takes place at different temperatures in the sample.
calcium monosulfoaluminate hydrate, 3CaO
Al2O3. CaSO4 The parallel observation of these two analyses (the peaks
xH2O, (JPCDS card no. 18-275), and tetracalcium of the DTA correspondent to the larger decays of TGA)
monocarboaluminate hydrate, 3CaO
Al2O3. CaCO3 provides information on the different components present
xH2O. Portlandite and calcite are typical hydration prod- in the tested material.
ucts of cement, and tetracalciummonocarboaluminate In the present study, the DTA plot of the original soil
hydrate is formed at early age of cement hydration in pres- (Fig. 9a) shows three endothermic peaks at 571.5, 760.5
ence of limestone or dolomite, while the and 867.4 °C, respectively, for the first representative of
tetracalciummonosulfoaluminate hydrate is a result of the the displacive and polymorphic transformation of
transformation of ettringite. All these components indicate the quartz, the second and third coming from the
a meaningful presence of cement.
The comparison with the material created with double
fluid jet grouting (Fig. 7c) soil reveals a smaller presence
of calcite, while most of the other phases produced by
cement hydration are negligible or totally missing.
Another relative effect seen from the XRD test on the
materials created with single and double fluid jet grouting
can be seen from the results in Fig. 8. Here the peaks are
at 26.7°, the diffraction angle typical of quartz are plotted
for the materials created with single and double fluid jet
grouting. The higher intensity observed for the double fluid
material confirms a higher relative presence of quartz min-
erals than the hydration products of cement.
The original soil and the materials formed by the two
injection systems were then subjected to Thermo-
Gravimetric (TGA) and Differential Thermal Analyses Fig. 8. Main XRD peak of quartz for the soil treated with single and
(DTA) (Fig. 9). The former test consists in measuring the double jet grouting.
 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954 949

Fig. 9. DTA and TGA of natural soil (a) and of the soil treated with single (b) and double (c) jet grouting.

decomposition of dolomite (MacKenzie, 1970). The same during its thermal decomposition, typically occurring at
test on the material treated with single fluid jet grouting temperatures between 150 and 450 °C. The CSH phase,
also shows peaks at 574, 700 and 811 °C, indicative of not easy to be distinguished with XRD because of its very
the transformation of quartz and dolomite. The transfor- low crystallinity (Hewlett, 2003), is the principal binding
mation of dolomite at lower temperatures than in agent in cement chemistry (Taylor, 1997). Together with
untreated soil can be explained by the reduction of the the former effect, there is another evident decay at 501 °
dolomite crystal size induced by the reaction with cement. C, indicative of the transformation of portlandite. Again,
In addition to the above peaks, the single fluid material this decay is totally absent in the double fluid material
shows other less intense endothermic peaks at 235 and (Fig. 9c), which is as a proof of the limited presence CSH
501 °C (Fig. 9b), the former (235 °C) indicative of calcium and portlandite.
aluminate hydrate phases, the latter (501 °C) induced by In conclusion, all the above analyses confirm a much
the thermal decomposition of portlandite. On the other lower presence of cement products in the column treated
hand, the DTA experiments performed on the double fluid with double fluid, possibly due to the diffusion of cement
material (Fig. 9c) reveal only two endothermic peaks, at in larger volumes, that could explain the lower strength
572.5 and 787.1 °C, related with the presence of quartz observed for this material from the uniaxial compressive
and dolomite. The peaks previously observed on the single tests.
fluid material at 235 and 501 °C, resulting from the decom-
position of calcium aluminate hydrate phases and port- 2.2. Case study #2: Clayey soils
landite have disappeared.
The curves obtained from TGA tests, reported in the The second field trial was performed in northern Italy,
same plot, show for the natural soil (Fig. 9a) a continuous in a layer of fine-grained material (Fig. 10a). The interpre-
weight loss of about 2% up to the room temperature up to tation of cone penetration tests with the Robertson (1990)
700 °C and two more marked weight losses (at 760 and chart (Fig. 10a) leads to the conclusion that the subsoil
867 °C) related to the decomposition of MgCO3 and consists of clayey and silty materials with variable propor-
CaCO3 present in the dolomite. As expected, no change tions. Two columns were then purposely created here, one
in the weight loss results at 571 °C from the polymorphic with single fluid adopting a cement-water ratio equal to
transformation of quartz. The same test performed on 0.71, the other with double fluid system adopting a
the soil treated with single fluid jet grouting (Fig. 9b) cement/water ratio equal to 0.95. Unlike the previous case
reveals a continuous weight loss of 4% for room tempera- study, where the cement content in the single and double
tures lower than 400 °C. This decay, significantly larger fluid columns was not specifically controlled, different com-
than the one seen on the untreated soil, comes from the position of grout and injection parameters were adopted in
dehydration of the calcium silicate hydrate phase (CSH) this second case, in an effort to obtain more similar cement
950 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954

Fig. 10. Field trial in clayey soil (a. CPTU profile and Robertson (1990) chart; b. coring of the sample; c. tested samples and laboratory equipment).

contents for the two columns (see Table 4). In fact, consid- and receivers placed at various locations on the sample’s
ering the diameter of the two columns, the cement content surface. The complete list of results from these tests,
estimated for double fluid (385 kg/m3) is slightly higher reported in Table 5, shows that the material created with
than for single fluid (353 kg/m3). As for the previous case single fluid jet grouting is much stiffer than the one created
study, this calculation has been made assuming an initial with the double fluid system. The same conclusion can be
porosity of the soil equal to 0.5 and similar composition derived from the uniaxial compression strength. In all
for jet grouted material and spoil. cases, the material strength is particularly low, but while
After injection, the two columns were excavated in their the mean qu for double fluid is about 150 kPa, the mean
top meter, samples were cored manually in the fresh jet strength of single fluid material is 500 kPa (333% higher).
grouted material (Fig. 10b) and taken to the same labora- It is worth noting that this difference cannot be attributed
tory used in the previous case study (Fig. 10c). to the different density of the two materials, even though
Before performing the uniaxial compression tests, the the dry unit weight of the samples formed with double fluid
samples were subjected to a series of non-invasive ultra- jet grouting is consistently larger than for the single fluid
sonic tests consisting in the measurement of the propaga- samples. Neither can the water-cement ratios of the mate-
tion velocities of compressional waves between sources rial be invoked to explain the noticed difference, since this

Table 4
Injection parameters and fundamental properties of the different columns of the field trial in clayey soils.
Column Nr and diameter Grout Grout Lifting Average Cement weight per Cement-soil Water/ Average uniaxial
[mm] of nozzles pressure flow rate speed diameter unit volume ratio cement compressive
of column ratio strength
# [MPa] [m3/s] [m/s] [m] [kg/m3] [MPa]
Single fluid 2x3.5 40 0.0045 0.01 0.6 353 0.71 1.95 0.5
Double fluid 2x3.5 40 0.0044 0.006 0.9 385 0.63 1.66 0.15
 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954 951

Table 5
Results obtained from the different laboratory tests on samples from the field trial in clayey soil.
Injection system Sample ID Dry unit Mean compressional UCS (kPa)
weight (kN/m3) wave velocity (m/s)
Single M#1-1 13.3 1055 427
M#1-2 12.9 961 484
M#2-1 13.1 1015 569
M#2-2 13.0 1222 –
Double B#1-2 14.6 490 –
B#1-3 16.4 370 175
B#2-2 15.0 343 122
B#2-3 15.5 – –

quantity is larger for the single than for double fluid mate- liquid and air) are present in the double fluid system (soil
rial. The cement content related to the soil present in the particles, liquid cement suspension and injected air), three
material is slightly larger for single than for double fluid interfaces should be expected (solid-liquid, solid-air and
and this could play some role in principle, though it would liquid-air). The increase in the number of interfaces and,
not be in the same direction as that suggested by Lee et al. particularly the additional presence of the liquid-air and
(2005), who noticed an opposite dependency. The question solid-air interfaces, can justify a looser distribution of the
needs to be investigated in more detail by looking at the binder above the soil particles.
composition of the material. Additionally, the carbonation rate of the portlandite
produced by the hydration of cement might play a role.
2.2.1. Mineralogical composition It is well known that a slower carbonation rate results in
The SEM analysis at high magnification reveals a mean- the growth of bigger calcite crystals in the conglomerate.
ingful morphological difference between the sample treated A faster crystallization rate, like the one induced by the
with single (Fig. 11a) and double fluid system (Fig. 11b). presence of air, may induce the formation of micro tensions
The two samples show similar particles with dimensions resulting in smaller crystals and worsen the mechanical per-
of a few microns, representative of the natural soil. How- formance of the conglomerate. The insufflation of air car-
ever, while the single fluid material shows the co- ried out in the double fluid system could then have
existence of sub-microparticles attached to the surfaces of triggered the faster carbonation of the portlandite, and this
the bigger soil particles, these sub-elements are absent in may explain the reduced strength of the material.
the double fluid material. These products, coming possibly Again, XRD analyses were performed on the different
from the hydration of cement, could be responsible for the materials, as samples dehydrated in air at 60 °C for 48 h.
greater resistance of the single fluid material. The morpho- Fig. 12 shows that single and double fluid treated materials
logical difference between the two materials, basically present peaks at similar diffraction angles typical of quartz
occurring from a more homogeneous distribution of the (q), calcite and clayey components like muscovite and clin-
hydration products of cement in the single fluid material, ochlore. Traces of feldspar and amorphous phases are also
can be justified by a different wettability by the cement sus- found in both materials. Therefore, the differences between
pension with the soil particles. In fact, for the single fluid the two materials is negligible with regard to the con-
system, a direct interaction can be assumed between soil stituent minerals present in the two cases.
particles and cement suspension and hence a single solid- Also, simultaneous DTA and TGA (Fig. 13) tests
liquid interface. However, because three phases (solid, carried out on the two materials show negligible

Fig. 11. SEM micrographs of the clayey soil treated with single (a) and double (b) fluid jet grouting.
952 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954

Fig. 12. XRD patterns of the clayey soil treated with single (a) and double (b) fluid jet grouting. (q = quartz; * = clinochlore, (Mg5Al)(SiAl)4O10(OH)8;
C = calcite, CaCO3; Mu = muscovite).

Fig. 13. DTA and TGA of the clayey soil treated with single (a) and double (b) fluid jet grouting.

differences. The DTA curves shows endothermic peaks cor- holic solution of phenolphthalein. When the pH values
responding respectively to the polymorphic transformation are lower than 8.2, the painted surfaces show no visible
of the quartz (at 578° and 575° for single and double fluid changes of colour, but for pH larger than 9.8, they assume
case) and to the decomposition of calcium carbonate (at a violet pigmentation (Collepardi et al. (2014)). This test
836° and 841° for single and double fluid case). The TGA showed a meaningful difference between the two materials:
curves also show equal weight losses in the range between the samples treated with single fluid assumed the typical
the room temperature and 900 °C, most likely due to the violet coloration (Fig. 14a), while those treated with double
decomposition of calcite. fluid remained unchanged (Fig. 14b). This different
To further clarify the implications of the above results, response of the clay treated with single and double fluid
another experiment was carried out on samples of single can be thus justified with the different degree of carbona-
and double fluid materials. After removing the outer envel- tion of the portlandite, Ca(OH)2, one of the cement hydra-
ope, two samples extracted from the columns were cut and tion products. While portlandite is characteristically
their intact middle surfaces were painted with a 1% alco- alkaline, when interacting with CO2, calcite (CaCO3) is

Fig. 14. Results of phenolphthalein test for soil treated with single (a) and double (b) fluid jet grouting and for soil treated with single fluid after 24 h of air
exposure (c).
 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954 953

formed, which has lower alkaline characteristics. The lack washing of the cement grout from soil particles surfaces)
of violet coloration for the double fluid sample could thus cannot not be excluded. In conclusion, the injection param-
stem from the carbonation of the portlandite activated by eters (primarily cement-water ratio and lifting speed of the
the injection of air, whereas the violet coloration of the sin- monitor) should be assigned according to the expected
gle fluid sample implies the presence of not yet carbonated diameter of the column obtainable in each case and by
portlandite. As further proof, the single fluid sample defining a suitable proportion of cement in the obtained
exposed to air after painting with phenolphthalein tended material.
to lose its violet coloration in a few hours (Fig. 14c). This To avoid this difference in the field trial performed in
change can be then ascribed to the carbonation of the small clayey soil, a larger amount of cement was injected in
amount of portlandite still present in the single fluid mate- the double than in the single fluid system to compensate
rial. In conclusion, the faster carbonation of the double for the larger dimension of column in the first case. The
fluid material induced by the insufflation of air rich with estimated amount of cement for unit volume of column
CO2 could have avoided the formation of bigger calcite is thus slightly larger for double than for single fluid. This
crystals and thus may be the reason for the lower strength result is also confirmed by the mineralogical experiments,
seen in the laboratory tests (Table 4). In fact, it is well that reveal a similarly low content of cement in both cases.
known (Jaewd et al., 1978; Ghosh et al., 2002) that cement The lower amount of cement than that in the first field
products are particularly sensitive to the presence of alka- trial can be explained by the difficulty in mixing grout
line carbonates as they accelerate the hydration of cement. and clayey soils, and also to the lower content of port-
In this case, these compounds arise from the contact landite. Portlandite, Ca(OH)2, one of the main hydration
between the newly formed portlandite and cement contain- products of cement, is generally less evident since its
ing alkali products. The hydration process determined by Ca++ ions deriving from the water dissolution tend to be
the carbon dioxide results in poorer mechanical perfor- exchanged with the cations of the clayey soil. However,
mance than conglomerates cured in absence of CO2. (e.g. in spite of their similar compositions, the strength of the
Kakali et al., 2000). double fluid is again much lower than that of the single
fluid material. The faster carbonation of the portlandite
3. Conclusions enhanced by the carbon dioxide (CO2) insufflated with
the air jet may have attributed to this result. This phe-
The present study focuses on the strength of the material nomenon induces the formation of micro tensions that
obtained with jet grouting to investigate the effects of two reduce the dimension of calcite (CaCO3) crystals. As a
very common versions of the technology: the single and result, while submicrometric crystals are seen from the
double fluid systems. Two real scale experiments were car- SEM analyses attached to the clayey particles treated with
ried out on site, with single and double fluid jet grouting single fluid jet grouting, the same effect disappears in the
performed in both sandy and clayey soil deposits. Several double fluid case.
samples were cored from the columns and subjected to In conclusion, the mechanical and mineralogical tests
mechanical tests and to different analytical procedures to performed in the two case studies converge to identify a
compare the composition of the material formed with the strong relation between injection system and material
two injection systems. The evidence shows that strength strength for jet grouting. The differences can be partly
is significantly lower for clayey than for sandy soil. This explained by the amount of injected cement in relation to
is due the more difficult erosive capacity of the jet in plastic the dimension of columns. Other reasons could be found
soils that originates because of incomplete mixing between in the different physical or chemical phenomena activated
the binding agent and original soil particles. by the injection systems, i.e. a possible segregation between
With a focus on the differences between the single and soil and grout promoted by the air jet, or in the different
double fluid system, the two studied cases show the positive carbonation rate of the portlandite due to CO2 present in
role of the air used in the double fluid system to increase the air jet.
the column diameter. As a counterpart, the uniaxial com- The above conclusions, certainly not exhaustive, should
pression strength is always much lower for the double fluid be examined further in other contexts, i.e. for different soil
than for the single fluid material. types, quantities and types of cement, injection systems
The analysis performed for the sandy soils identifies the etc., to understand the relevant mechanisms activated by
proportion of cement present in the material as the main jet grouting. This kind of analysis cannot be performed
cause of the observed difference in strength. Uniaxial com- uniquely in the laboratory because factors related to the
pressive strength was found to increase with the cement/ execution of the technology cannot be reproduced.
soil ratio of the material. The lower presence of cement Although affected by a larger complexity and uncertainty
in the double with respect to the single fluid columns was in the interpretation of results, site investigations like the
confirmed by SEM, XRD, TGA and DTA analyses, which ones herein carried out are the only alternative. The results
were carried out to observe the mineralogical composition of this investigation represents an opportunity to improve
of the materials. The large differences observed from these jet grouting technology, optimize its mechanical perfor-
tests suggest that other negative effects of the air jet (e.g. mances and lead to more rewarding applications.
954 G. Modoni et al. / Soils and Foundations 59 (2019) 942–954

References Lee, F.H., Lee, Y., Chew, S.H., Yong, K.Y., 2005. Strength and modulus
of marine clay-cement mixes. J. Geotech. Geoenviron. Eng. 131 (2),
Atangana Njock, P.G., Shen, J.S., Modoni, G., Arulrajah, A., 2018a. 178–186.
Recent advances in horizontal jet grouting (HJG). Arab. J. Sci. Eng. Mackenzie, R.C., 1970. Differential Thermal Analysis vol. 1, 479.
43, 1543–1560. Mascolo, G., Mascolo, M.C., Vitale, A., Marino, O., 2010. Microstructure
Atangana Njock, P.G., Chen, J., Modoni, G., Arulrajah, A., Kim, J.H., evolution of lime putty upon aging. J. Crystal Growth 312, 2363–2368.
2018b. A review of jet grouting practice and development. Arab. J. Miki, G., 1973. Chemical stabilization of sandy soils by grouting in Japan.
Geosci., 11–459 Proceedings of the 8th ICSMFE, pp. 395–405.
Coats, A.W., Redfern, J.P., 1963. Thermogravimetric analysis: a review. Modoni, G., Croce, P., Mongiovı̀, L., 2006. Theoretical modelling of jet
Analyst 88–1053, 906–924. https://doi.org/10.1039/AN9638800906. grouting. Géotechnique 56 (5), 335–347.
Collepardi, M., Collepardi, S., Troli, R., 2014, Il nuovo calcestruzzo, Enco Modoni, G., Bzówka, J., 2012. Analysis of foundations reinforced with jet
Srl. grouting. J. Geotech. Geoenviron. Eng. ASCE 138 (12), 1442–1454.
Croce, P., Flora, A., Modoni, G., 2014. Jet grouting: Technology, Design Modoni, G., Flora, A., Lirer, S., Ochmański, M., Croce, P., 2016a. Design
and Control. Taylor & Francis Group, ISBN 978-0-415-52640-1. of jet grouted excavation bottom plugs. J. Geotech. Geoenviron. Eng.:
Croce, P., Modoni, G., 2007. Design of Jet Grouting Cut-offs, Ground ASCE 142 (7).
Improvement, Thomas Telford, 11-1, pp. 11–19. Modoni, G., Wanik, L., Giovinco, G., Bzòwka, J., Leopardi, A., 2016b.
EN 197-1, 2011, Cement – Part 1: Composition, specifications and Numerical analysis of submerged flows for jet grouting. Proc. ICE,
conformity criteria for common cements, Technical Committee CEN/ Ground Improvem. 169 (1), 42–53.
TC 51, Cement and building limes. Ochmański, M., Modoni, G., Bzòwka, J., 2015a. Numerical analysis of
Eramo, N., Modoni, G., Arroyo Alvarez de Toledo, M., 2012. Design tunnelling with jet-grouted canopy. Soils Found. 55, 929–942.
control and monitoring of a jet grouted excavation bottom plug. In: Ochmański, M., Modoni, G., Bzòwka, J., 2015b. Prediction of the
Viggiani (Ed.), Proc. of the 7th Int. Symposium on Geotechnical diameter of jet grouting columns with artificial neural networks. Soils
Aspects of Underground Construction in Soft Ground, TC28 IS Found. 55–2, 425–436.
Rome. Taylor & Francis Group, London, 16–18 May 2011, pp. 611– Qiu, J., Liu, H., Lai, J., Lai, H., 2018. Investigating the long-term
618. settlement of a tunnel built over improved loessial foundation soil
Flora, A., Modoni, G., Lirer, S., Croce, P., 2013. The diameter of single, using jet grouting technique. J. Perform. Constructed Facil., 32–35
double and triple fluid jet grouting columns: prediction method and Reimer, L., 1998. Scanning Electron Microscopy – Physics of Image
field trial results. Géotechnique 63–11, 934–945. Formation and Microanalysis. Springer.
Flora, A., Modoni, G., Croce, P., Siepi, M., Kummerer, C., 2017. What Robertson, P.K., 1990. Soil classification using the cone penetration test.
future for jet grouting? A European perspective. In: Proc. of the Int. Can. Geotech. J. 27–1, 151–158.
Conf. Grouting 2017. ASCE library, Honolulu, July 2017, pp. 358– Salvatore, E., Modoni, G., Andò, E., Albano, M., Viggiani, G., 2017.
381. Determination of the critical state of granular materials with triaxial
Ghosh, S.N., 2002. Advances in Cement Technology: Chemistry, Man- tests. Soils Found. 57–5, 733–744.
ufacture and Testing, second ed. Tech. Books International, New Shen, S., Wang, Z., Yang, J., Ho, C., 2013. Generalized approach for
Delhi, India, pp. 687–697. prediction of jet grout column diameter. J. Geotech. Geo-Environ.
Hewlett, P., 2003. Lea’s Chemistry of Cement and Concrete. Butterworth- Eng.: ASCE.
Heinemann. Taylor, H.F.W., 1997. Cement Chemistry. Thomas Telford, p. 548.
ICDD, International Centre of Diffraction Data, 2017, http://www. Tinoco, J., Gomes, Correia A., Cortez, P., 2014. Support vector machines
icdd.com/. applied to uniaxial compressive strength prediction of jet grouting
Iolli, S., Modoni, G., Chiaro, G., Salvatore, E., 2015. Predictive correlations columns. Comput. Geotech., 132–140
for the compaction of clean sands. Transp. Geotech. 4, 38–49. Toraldo, C., Modoni, G., Ochmański, M., Croce, P., 2018. The
Jaewd, I., Skalny, J., 1978. Alkalies in cement: a review and effects of characteristic strength of jet-grouted material. Géotechnique 68 (3),
alkalies on hydration and performance of Portland cement. Cem. 262–279.
Concr. Res. 8 (1), 37–51. Van der Stoel, A.E.C., 2001. Grouting for Pile Foundation Improvement
Kakali, G., Tsivilis, S., Aggeli, E., Bati, M., 2000. Hydration products of PhD thesis. Delft University Press, p. 217.
C3A, C3S and portland cement in the presence of CaCO3. Cem. Wanik, L., 2017. Influence Analysis of Selected Parameters on Geomet-
Concr. Res. 30 (7), 1073–1077. rical and Mechanical Properties of Jet Grouting Columns PhD thesis.
Kaushinger, J.L., Perry, E.B., Hankour, R., 1992. Jet grouting: state of the Silesian University of Technology, Gliwice (Poland).
practice. In: Proc. Grouting, Soil Improvement and Geosynthetics, Wanik, L., Mascolo, M.C., Bzόwka, J., Modoni, G., Shen, J.S.L., 2017.
vol.1. ASCE, New York, pp. 169–181. Experimental evidence on the strength of soil treated with single and
Kitazume, M., Terashi, M., 2013. The Deep Mixing Method. CRC Press, double fluid jet grouting. In: Proc. of the Int. Conf. Grouting 2017.
Taylor & Francis group, p. 405. ASCE library, Honolulu (USA), pp. 52–61.
Klug, H.P., Alexander, L.E., 1974. X-Ray Diffraction Procedures: For
Polycrystalline and Amorphous Materials. Wiley.